key: cord-0940883-oei7dr4e
authors: Randall, Pierce; Bernstein, Justin
title: Reciprocity and the ethics of giving during pandemics
date: 2021-03-19
journal: J Soc Philos
DOI: 10.1111/josp.12408
sha: 6f65d1c1e0b5b87fcddd3767c65f74a7a50c525c
doc_id: 940883
cord_uid: oei7dr4e

nan

The COVID-19 pandemic has had devastating effects worldwide. In response, many individuals have made significant sacrifices. These sacrifices include those made by essential workers who risk infection in order to ensure that others can have access to adequate food, medical care, electricity, and other goods. Individuals are also asked to sacrifice through social distancing. Social distancing brings about a collective benefit, in the form of a reduced overall transmission rate of coronavirus. Yet it is also quite burdensome. Many workers have foregone earning incomes, businesses have ceased operating, and children have missed educational opportunities, to name just a few prominent costs. In response to the sacrifices people have borne during this pandemic, a natural question concerns what you, as an individual, can do to help. Our answer in this paper is that individuals who have not been significantly burdened by the collective response to the pandemic have an obligation to give to those who haveespecially those who have been burdened due to background injustices. We focus on the American context in particular, although the arguments could be modified to apply to other contexts.

In Section 2, we consider two different rationales for donating to people who have borne especially significant hardships during the pandemic. While one approach would be to focus on duties of aid as a rationale for giving, we argue that a reciprocity-based approach is more sensitive to the ways in which people undertake burdens in order to produce a collective good. In Section 3, we defend the claim that compensation is owed to those who have been unfairly burdened in order to produce a collective benefit. In Section 4, we address who has been unfairly burdened. We endorse an anti-perversity condition: those who are worst-off because of background injustice have a stronger claim to compensation than, for example, wealthy investors who have seen a decline in the values of their portfolios. In Section 5, we argue that giving money to those who have borne unfair burdens in response to the COVID-19 pandemic can be understood as a way of addressing a predictable epistemic limitation on the part of the state. In Section 6, we consider an objection from those who claim that we should be giving money to those in most need around the world rather than locally.

Pierce Randall and Justin Bernstein have contributed equally to this paper.

water, each resident is required to refrain from watering their lawns or their plants. The community largely complies with this rule and, as a result, the community's common supply of potable water is not depleted. In this context, it would be morally objectionable for you to free-ride by watering your lawn. Your violation of the water usage restriction would not be objectionable because your act, taken by itself, harmed the community by causing the depletion of its potable water. Instead, it would be objectionable because you would be taking advantage of the cooperative sacrifices voluntarily undertaken by others by enjoying the benefits of access to potable water without doing your fair share. In the context of the pandemic, there is a clear sort of analog. Social distancing during a pandemic is like observing watering restrictions during a drought. We frequently hear public health experts and influencers in media enjoin us to "do our part" by social distancing and thereby contribute to the effort to ensure that we all enjoy the benefits of reduced transmission rates and a flattened curve. 1 According to this rationale, those who are able but refuse to socially distance might very well end up enjoying those benefits, but they will not have done their fair share-they are guilty of free-riding.

To return to the drought example, imagine next that some members of the community are botanists who need to water plants in order to make a living whereas others just enjoy having a beautiful lawn. All else being equal, it seems unfairly burdensome for the botanists if they were to have to forego watering plants. And if we accept that these burdens are unfair, there are a few intuitive types of responses. The first would be to relax water restrictions and thereby allow everyone to water the plants. This response would not unfairly burden botanists, but it would lead to a depleted water supply. The second kind of response would involve trying to carve out an exemption to the rule so as to ensure it was not unduly burdensome; the botanists might receive an exemption that permits them to water more than the rest of the community. While watering privileges would then be distributed unequally, they would not obviously be distributed unfairly. Indeed, an equal distribution of watering privileges would be especially burdensome for botanists, and, perhaps would be unfair as a result. The third kind of response would involve compensating the botanists. The other members of the community could start a fund to pay for the loss of revenue the botanist would experience from compliance with watering restrictions.

The analogy we wish to draw is probably apparent. In our current context, we have been enjoined to socially distance to promote various goods for the community. Yet for some people, especially those who have to forego educational or economic opportunities, these costs are not equal. Indeed, in some cases, we might think that such costs are unfair; some members of the community bear unfair burdens for the sake of producing a benefit for the community as a whole. A similar point applies to the distribution of benefits from promoting the community's good. If some people benefit much more than others when it comes to the production of a public good, then our sense of "fair shares" will also be affected. So, for instance, ensuring a river remains clean will benefit those who make their living fishing from that river more than those who do not get their water from that supply and do not need to fish. Similarly, reducing transmission rates will benefit those especially at risk from the virus or those who are more likely to need emergency services when there is a shortage of hospital beds. COVID-19 has a much higher case fatality rate for those from older age groups than younger age groups. Children are far less likely to die from COVID-19 than those fifty and older. If we were to consider those benefits alone-independently of various indirect benefits children enjoy in virtue of reduced transmission rates-then older generations stand to benefit more than young children from social distancing measures. In other words, children take on significant burdens in the form of school closures. Yet they enjoy the direct benefits of social distancing far less than other members of our society.

A similar point applies to those who take on exceptional risks for the sake of the common good by continuing to work. In these cases, workers remain employed and paid for the services that they provide; the feeling is that perhaps they are not compensated enough given the extraordinary circumstances they find themselves in, that they are entitled to some sort of hazard pay. People now facing additional risks normally rely on markets or public spending to set their wages. The rationale for giving to them is to make up for deficiencies in this process in public health emergencies. And this rationale resonates with many people; there is a call to give these workers their "fair share" (Allen 2020) or to "take care" of those who have "taken care of us" (Stewart 2020) . More concretely, we are regularly enjoined to give larger tips to service workers or assist medical workers who get sick or experience financial trouble relocating to areas where they are most needed.

The upshot of this discussion is that on many views about the fair distribution of burdens and benefits, our collective response to the pandemic has involved some bearing unfair costs for the sake of the good of the community as a whole. Different people could arrive at this conclusion despite disagreeing about the exact nature of the burdens and benefits of this collective response. For instance, some might draw on resourcist views and construe the burdens of social distancing solely in terms of lost economic or educational opportunities (e.g., Daniels 2008; Rawls 1999b) . Others might spell out the burdens in terms of loss of preference-satisfaction or the capabilities approach (Robeyns 2016) . Similarly, proponents of different allocative principles would converge in their judgment that the burdens and benefits of social distancing are distributed unfairly. Regardless of one's preferred view, many will share the judgment that social distancing has been unfairly burdensome on some members of our society-although more needs to be said about who has been unfairly burdened in particular, an issue we return to below.

Given that social distancing measures distribute benefits and burdens unfairly, what might we do? Well, as we saw from the discussion of the water restrictions example, three especially intuitive responses emerge when a rule provides a benefit to a community but is unfair. First, some might argue that given the unfairness involved, we should seriously consider unshuttering businesses, opening schools, and relax other policies that have been implemented in order to mitigate the pandemic. We will assume such a response is infeasible, at least at the time this paper has been written. Most guidelines for "reopening" society recommend doing so only after at least two weeks of declining caseswhereas others suggest months of distancing before reopening (Emanuel et al. 2020; Gottlieb et al. 2020 ). Most states have not reached the relevant threshold to satisfy those recommendations. And even if ultimately unfair, we will assume in this paper that the terrible epidemiological consequences of reopening count against doing so; fairness is not the only value at stake.

The second way to make the burdens and benefits of social distancing fairer would involve issuing exemptions. Those for whom social distancing is especially burdensome would be permitted to, say, return to work or school, while those for whom it is not burdensome would need to continue to stay at home. We will proceed as though this option is also infeasible for similar reasons. If the especially burdened members of society were permitted to return to work, we should worry that this would forego the epidemiological advantages of social distancing. Moreover, reopening society, either for all of society or for the unfairly burdened individuals, would not compensate them for the sacrifices they have already made; such policies are ultimately "forward-looking," but many individuals have already made the relevant unfairly burdensome sacrifices.

A third possibility would be to compensate those who have had to bear unfair burdens in order to promote the common good. Because social distancing measures have been especially burdensome for certain individuals, we might think they are owed compensation from those of us who have not borne comparable burden-those of us who have continued to telework or enjoy educational opportunities but nonetheless benefited from the major sacrifices of others. In effect, compensation might shift some of the burdens onto those who have benefited without bearing comparable burdens, themselves. Moreover, it would do so without compromising the epidemiological gains of our collective response. Compensation, then, would seem to be an appealing option; it would mitigate the unfairness involved in our collective efforts to reduce overall transmission rates and flatten the curve, but it would do so without foregoing the epidemiological benefits.

To recapitulate: the argument has been that when thinking about giving away money in the current pandemic, it would be incomplete if we were to merely consider the needs that arise due to the pandemic itself. One should also consider the burdens and benefits that arise from our collective response to the pandemic. Many Americans have engaged in a collective effort to try to reduce the rate of transmission for our entire community-we have radically altered our way of life to protect all of us. Yet the costs and benefits from this radical change have not been distributed fairly. A natural thought, then, is that compensation is owed to those who have borne unfair burdens or have benefited far less than others in the community.

Who bears fair or unfair burdens during the pandemic? One view might be that those who have lost the most in monetary terms due to our collective response are those who are entitled to compensation. Yet this seems implausible for two reasons. The first is that who has experienced the greatest loss in monetary terms does not necessarily correspond to who has borne the greatest burdens. A poll tax of $1000 for each citizen or resident will be more burdensome for the least well-off members of society than the wealthiest, even though the wealthiest members of society are taxed the same amount. Similarly, even though Shake Shack might suffer larger financial losses in absolute terms than smaller restaurants during this time, this does not mean that CEOs or shareholders in Shake Shack have been burdened more than owners of smaller, non-franchised restaurants. This in part explains the condemnation the company received for accepting loans intended to support small businesses through the Paycheck Protection Program before agreeing to give them back (Yaffe-Bellany 2020).

The second point is that when assessing the benefits and burdens of some policy or set of policies, not all burdens have the same source. In particular, some burdens are due injustice, and should not be assessed the same as other kinds of burdens. To illustrate with an example, imagine that, due to existing injustices, someone is born in an environment in which she is more exposed to air pollution than other members of the community. She develops asthma as a consequence. Imagine further that in her society, there is universal health insurance although people have to pay higher premiums for pre-existing conditions. Insisting that this person should pay a higher premium is not merely unfair. It is morally perverse; she is being required to bear a particular burden precisely because she is a victim of injustice. 2 This example is meant to highlight a certain kind of consideration that should inform assessments of whether a distribution of burdens and benefits is fair. Namely, it would be perverse to claim that individuals are obligated, as a matter of fairness, to take on certain burdens if those burdens arise due to past or ongoing injustices. This applies to the present context because individuals from low-income backgrounds, African Americans, Latinx individuals, and Native peoples have all been especially burdened. 3 Individuals from these social, racial, and ethnic groups disproportionately lack access to the kinds of jobs that allow for teleworking and thus are disproportionately unemployed or face great risk as essential workers; they disproportionately are deprived educational opportunities that enable access to those jobs; and distance learning is especially burdensome for children from these groupsdue to, say, the lack of access to laptops or constraints on parents. In short, these individuals have been especially burdened because of historical and ongoing injustices.

Importantly, individuals from these social, racial, and ethnic groups are also disproportionately suffering the effects of the pandemic, itself. So, the collective response to the pandemic, while especially burdensome for these individuals, stands to benefit them a great deal; indeed, some object to ending social distancing measures precisely because they worry these disadvantaged individuals will be especially harmed by the consequent rise in transmission of the virus (Serwer 2020) . A natural objection to the argument so far, then, is that many of these individuals are the primary beneficiaries of our collective response to the pandemic. They might be more burdened than others, but they also benefit more. So, perhaps, the benefits and burdens of our collective responses are distributed fairly, after all, and considerations of reciprocity do not entitle these individuals to compensation.

In reply, an initial point is that part of the reason why individuals from these backgrounds have higher rates of COVID-19 is because they work essential jobs where the risk of exposure is higher than, say, for those who telework (Hawkins 2020) . As discussed, some of these individuals are already taking on burdens for the common good in the form of heightened risk of contracting the diseaserather than being individuals who stand to benefit especially from social distancing.

Second, and more fundamentally, the concern about moral perversity applies not only to burdens but also to benefits. The reason why individuals from these groups have died in higher numbers is precisely due to salient background injustices-such as being disproportionately employed as essential workers at heightened risk, not being able to take time off from work, enjoying inadequate medical care-sometimes due to prejudice or stigma-or living in areas that are more likely to cause certain health conditions, such as asthma or hypertension, that exacerbate the disease (Braveman et al. 2010; Owen, Carmona, and Pomeroy 2020; Shonkoff and Williams 2020; Wadhera, Wadhera, and Gaba 2020) . In other words, even if we grant that social distancing is, in absolute terms, very beneficial for members of these groups given the grave threat the novel coronavirus poses to them, this is due in large part to background injustices.

What is the upshot? When determining whether the benefits and burdens that accompany social distancing are fair, we should take into account burdens relative to one's overall resources or opportunities. This counts against just looking at burdens or benefits in absolute terms. We should also look at the source of burdens and benefits. Some individuals who shoulder especially large burdens during this time do so because of persistent injustices. To insist that the losses those individuals experience amount to their "fair share" would not merely be false; it would be perverse.

Accordingly, when we think about whether the distribution and burdens and benefits accompanying social distancing are fair, we should pay special attention to those from less advantaged backgrounds and are burdened due to injustice. Our claim in this section has been that many such individuals are doing more than their fair share in producing a benefit for the community as a whole. When combined with the conclusion of the previous section-that those of us who are not unfairly burdened have duties of fairness to compensate those who are unfairly burdened-we arrive at the main thesis of this paper: we are obligated to compensate those members of our community who come from less advantaged backgrounds, especially those who are burdened to historical and ongoing injustice.

In the previous three sections, we have argued that benefits and burdens of social distancing are unfairly distributed and that this unfair distribution ought to be remedied. But it does not follow from this argument that individuals who are relatively unburdened by social distancing have an obligation to help those who are disproportionately burdened (low-wage workers, school children) through direct assistance. One possibility is that institutions, such as government agencies, should be providing this kind of assistance, rather than individuals.

Government has an important role to play in helping to secure terms of social cooperation that are fair according to norms of reciprocity. One important function of government is to facilitate the production of collective goods through practices like social distancing by making it more difficult to free ride. One might cogently reason that one would benefit from others' efforts to socially distance, since this would reduce the overall transmission rate, without being willing to give anything up by social distancing oneself. States help keep this kind of free-riding in check by legally mandating that certain businesses close, as well as sanctioning individuals who violate social distancing rules.

Government also has a role in helping to alleviate the burdens some experience through social distancing. Many states have provided economic assistance to households that are likely to be hit the hardest by the response to the COVID-19 pandemic. In these cases, as with making free-riding more costly, the state is addressing a motivational problem on the part of private individuals: many would not help those who are disproportionately burdened by our collective response to the pandemic, while taking advantage of their sacrifices. The government's financing of assistance through taxation makes this assistance mandatory.

For example, the United States government has provided $1200 in assistance through the CARES Act to individuals earning under $75,000 a year, with a graduated reduction in benefits for those earning beyond that amount. While this policy has been referred to as a "stimulus," the policy is better understood as assistance, for two reasons. First, the government's aim has not been for consumers to rush out and purchase goods and services from businesses that are actually made vulnerable because of social distancing. Those businesses are closed or seeing less business precisely because of social distancing. Second, the assistance was means-tested, which is a good approach if the aim is to direct funds to those with lower incomes, but a bad approach if the aim was to get checks out sooner to stimulate spending. Instead, the CARES Act funds to households are better understood as a form of economic aid, while millions are facing financially precarious conditions. While assistance through the CARES Act has been targeted to some extent, this targeting has been quite rough-grained. This is because governments face a difficult epistemic problem when attempting to assess who is experiencing greater burdens because of social distancing. Not all employees who make similar wages are similarly burdened by social distancing. A graduate teaching assistant earning a PhD in economics is likely in a less dire position financially than a single-parent household earning the same income from a minimum-wage job. Some households experience higher debt than others or have difficult needs that the government cannot easily target. And, as we have mentioned, some workers are able to safely telework from home, while others experience layoffs, and some children have access to resources that make remote education far more feasible for them than for less advantaged children. When considering what obligations individuals have to assist others through giving money beyond the efforts by the state at correcting for the disproportionate burdens experienced by some from social distancing, it is important to note the limitations to what the state is capable of knowing about the hardships faced by particular individuals, and who would possess this kind of knowledge. In many cases, individuals are in a better epistemic position to recognize whether or not they are unduly burdened by social distancing than the state is, and so in a better position to know that economic assistance would be better targeted helping others.

The fact that the government has only a limited ability to predict who exactly is most burdened by social distancing and will need assistance is a deep problem. The relevant information includes not only predictions about individuals' economic behavior, but also about their preferences and expectations. To assess the degree to which someone is burdened by social distancing, the government would have to know whether working from home is an adequate substitute for their working in an office, whether they have other options if they are unable to work, to what degree they depend on having access to childcare, the tolerance for experiencing a greater risk of getting sick, and whether or not they have important life plans and milestones that are interrupted by having their options removed under social distancing. States face obstacles in gathering this kind of information. This is similar to why central planning tends to be an inefficient way to allocate scarce resources, since information about individuals' preferences and trade-offs are normally gathered through economic exchange (Hayek 1945) . A government's efforts to target financial assistance will necessarily be rough-grained, given its epistemic limitations about knowing the degree to which someone has been burdened by social distancing. It will predictably generate false positives, by giving to some people who have not seen financial hardship or a loss of income due to the crisis; and it will generate false negatives, because it fails to compensate some who are profoundly burdened by the policies.

Individuals generally do, however, have the relevant kind of information available regarding themselves. Many people are aware that they are not burdened by social distancing, either because it does not frustrate any opportunity that may wish to take or because they are not being asked to take on excessive risk in their daily lives. 4 Those who are not burdened can give to charitable organizations, which often possess local knowledge of what kinds of needs are being generated in a community (homelessness, food insecurity), the preferences community members, as well as expertise regarding how best to satisfy those needs and preferences (Kingma 1997) . Individuals are also in a better condition than the government to assess when giving direct cash transfers to neighbors, family members, or individuals through crowd-sourced donation websites like GoFundMe.

A worry about this point of the state's lack of information is that it would seem to apply to all instances of need-based assistance that the state applies. If the state is not particularly good at determining who is greatly burdened by the social response to COVID-19, then does it make sense to have federal poverty-relief programs, or assistance for those affected by hurricanes or floods? And in cases where such programs do exist, do individuals who benefit from them without being excessively burdened have an obligation to donate some of the money to those who have been? We provide two responses to these concerns.

The first is that social distancing raises a particularly difficult epistemic problem for the state with respect to compensating people in need. So, we should expect the state to misallocate relief assistance to a greater degree than with respect to other crises, and for more individuals to receive assistance without being disproportionately burdened. In the case of something like hurricane assistance, the state's epistemic challenges are more tractable. Government agencies can model how many people's homes will be destroyed in a given area, how long it will take to rebuild, and how much insurance market assistance would be needed to encourage rebuilding. For the most part, people's preferences to live in homes on the coast that are not destroyed by hurricanes do not change. Rather, a hurricane is an external shock to people's lives with a somewhat predictable, or at least modelable, cost. The economic burdens of social distancing, however, are more difficult to predict, because the economic damage is caused by changes in people's preferences and behaviors. The state will struggle to anticipate how much risk an individual household is willing to take on, or should be willing to take on, before going back to work. Since the disease is likely to persist for some time and we might be waiting for a vaccine for up to as many as 4 years, the state cannot predict how this will change people's preferences for where to work, what kinds of businesses to patronize, or what kind of communities they want to live in. Because people's preferences and behaviors are shaped in unpredictable ways by social distancing, the state cannot easily predict need on the basis of the crisis. Additionally, there is a significant lag in the data that the state can draw to assess need: it can use income reported in past tax returns to predict who will be neediest, but, obviously, people's financial conditions may have changed since and will be affected by the crisis.

The second response is just to bite the bullet about the state's ability to gather fine-grained information about the extent to which members of society are burdened by disasters or its own policies. States do not have an enviable epistemic position when assessing the degree of burden or need individuals have that would entitle them to compensation. This does not mean that the state should get out of the business of trying to alleviate burdens in the population more generally. It should simply be recognized that its attempts to do so will inevitably be less precise than would be ideal, sometimes helping those who are not particularly burdened and sometimes failing to adequately help those who are.

This raises a deeper point about the relationship of giving money and government assistance. At least sometimes, individuals have a moral duty to help see to it that others who share collective burdens with them are not unfairly burdened and that they are not living at the expense of others' contributions. We have argued that this is the case with respect to the collective response to the pandemic, but it is likely the case with many other kinds of collective burdens: wartime rationing, possibly conscription in a military conflict with a just cause, and forms of social insurance such as healthcare provision and retirement pensions. The state can help us discharge our duties to each other: it can help overcome free-rider problems, since it makes some transfers for the purpose of assisting others mandatory. But we should expect the state to be somewhat imperfect in assessing the needs of others or the extent to which they are burdened. The fact that the state is involved in the business of alleviating burdens does not obviate individuals' responsibilities to help others disproportionately burdened by social cooperation. Instead, individual charitable giving helps to solve the "last mile" problem of fulfilling our obligations of fairness in contribution to each other.

Someone who receives relief money because the state misallocated it to them, by failing to register that they are not particularly burdened by social distancing, has an obligation to see to it that the money goes to those who are disproportionately burdened. It is, in some sense, not their money to keep, since it is only the result of a mistake on the part of the state that they receive it at all. Indeed, some recipients of a stimulus check who have donated the money have expressly said they do not view the check as "their money" (Albrecht 2020). Our argument suggests one way to explain why these people are right: keeping the money would be unfair to those who are disproportionately burdened. Giving one's stimulus check to those who are disproportionately burdened amounts to helping the government do its job-by correcting for its mistakes.

We have argued that our duties of reciprocity to one another provide reasons to give charitable assistance to others in one's community who are shouldering the burdens of social distancing. But this raises two related sets of concerns. First, many philosophers have also argued that we have obligations to provide assistance to others grounded in the duty of rescue. For instance, Peter Singer (1972) has forcefully made the case that problems like global poverty and malaria are dire, that individuals in wealthy countries are able to save lives by giving money directly to alleviate these problems, and that individuals have an obligation to do so if they can without giving up something of significant moral worth. Even while social distancing and during a pandemic, most Americans are better off than those who experience severe poverty (the over one billion people living on less than $1.25 a day) or who live in malaria-stricken regions. It may seem that assisting them is more urgent than giving to other Americans. We need some explanation for how to weight these claims for assistance against the reciprocity-based case we have made for giving locally.

Second, some theorists of global justice have argued that cooperation is not limited by national borders (Moellendorf 2002; Tan 2004) . This may seem especially clear in a pandemic, where the coronavirus has crossed borders and has appeared in nearly every country in the world. So even absent the argument from the duty of rescue, it may seem that there are also people in low-income countries who are contributing to a collective benefit of reducing the spread of the coronavirus through social distancing. Insofar as many of these individuals also experience significant unmet material needs under even normal circumstances, it may seem that they are the most significantly burdened by the crisis, and so have a stronger claim to assistance than people who are burdened but who live in relatively affluent societies.

In this section, we offer four responses in reply to these concerns. The first two responses address the first concern, that there is a stronger case to give to assist those in dire poverty or facing systemic health burdens like malaria in low-income countries. This concern really amounts to whether duties of reciprocity should determine how we give during the pandemic, or whether duties of aid should.

First, while the duty to rescue may be particularly stringent, obligations grounded in reciprocity are correlative to the claims of others, and these claims constrain how we may fulfill duties of rescue. To be clear, this is not to say that obligations of reciprocity always override duties of rescue. Instead, it is only the weaker, but intuitive, position that claims of reciprocity, like claims of the recipient of a promise or the property-owner, ordinarily restrict how one may go about fulfilling one's duties of rescue. It is normally only permissible to discharge the duty of rescue with resources that are rightfully one's own, and which one has a claim to. People who have received state assistance in the form of a relief check, but who are not in great need of relief because they have not been significantly burdened by social distancing, received a benefit that others had a stronger claim to in virtue of being burdened by social distancing. It is only because the state made a mistake in helping us discharge our obligations of reciprocity to one another that accidental beneficiaries received the money in the first place. As discussed above, in some sense, the money from a relief check is not really theirs to decide what to do with.

To illustrate how claims of reciprocity constrain how one may discharge a duty of rescue, consider the following example. John owes Sally $100 and promised to pay her on Friday. It may be the case that John also has an obligation, grounded in the duty of rescue, to donate $100 to OXFAM. But it is less clear that he has an obligation, or even that it would be permissible, to donate the $100 he owes Sally. It is, after all, money Sally has a claim to, and that significantly constrains what John may permissibly do with it. Analogously, suppose that Zadie and Louise live together on a communal farm. They share resources in common. They pitch in when things are broken. They till each other's fields. Now, the tractor is broken, and it is Louise's turn to fix it so that she can harvest Zadie's corn on Friday. The parts will cost Louise $100, and it will take her a day of labor. As in the case with John, it may be that Louise has a general obligation, grounded on the duty of rescue, to donate $100 to OXFAM. But like in the promissory obligation case, it is not clear that she is permitted to spend the $100 that she needs to fix the tractor in order to discharge that duty. She owes it to Zadie, after all, to fix the tractor. Louise has this obligation because she is enmeshed in a form of social cooperation with Zadie to which a norm of reciprocity applies, like co-nationals who are making mutual sacrifices in order to lower the transmission rate of COVID-19. As this example illustrates, obligations grounded on reciprocity norms constrain the discretion one has in discharging a duty of rescue. 5 Second, while we have argued that obligations of reciprocity ordinarily constrain how we may fulfill obligations to rescue, we do not claim that the duty to rescue could never override the valid claims others may make on one's property or money. To use a well-traveled example, if you see someone drowning in a pond, and you are able to save them only if you ignore the "no unauthorized use" sign next to someone's rowboat and to use it, then you ought to use the rowboat. Yet using the rowboat infringes upon the owner's property rights, especially in light of their clearly expressed wishes. In this case, the obligation to rescue is strong, and the boundary crossing relatively minor, so infringing upon property rights in order to rescue the person drowning would be justified. The test for whether the rowboat owner has a valid claim against your action-that one has discharged their obligation of rescue only by something that, by right, was someone else's-is whether there is any residual moral claim on the part of the owner. In this case, there is. The rescuer at least owes the owner of the boat an apology and an explanation of what happened, and the owner is owed compensation for damages-perhaps by the rescuer or the person rescued (c.f., Feinberg 1978, 102) .

Similarly, while we do not take the position that obligation duty to rescue by helping those in dire poverty could not conceivably outweigh our obligations of reciprocity to help those burdened by social distancing, moral residue would also be left over in this case. Someone who receives a check but is not in great need unfairly received more than they should have. If they use this check to discharge a duty of rescue by donating it to OXFAM instead of helping to compensate those who have been unfairly burdened at home, then they owe those who have been unfairly burdened an apology, or they should work to make amends some other way. Simply assuring themselves that they did the right thing and spending their remaining discretionary income on themselves would not be permissible in this case.

The next two points are in response to the concern that the response to the pandemic should be understood as a global phenomenon. The third point is that while the pandemic is a global phenomenon, the primary means by which the global population has cooperated to address this problem, social distancing, is not. Since transmission rates of coronavirus do not correlate strongly by country (since most global transit is temporarily closed), social distancing in one country primarily benefits only the people in that country by helping to lower the transmission rate there. Americans who socially distance by being out of work are primarily helping to reduce the risk of the virus spreading to other Americans, rather than to people in other countries. That is, the benefits of social distancing are especially pronounced locally. So, it is well-motivated to focus on local benefits and burdens associated with the collective effort to mitigate the spread of the novel coronavirus at home.

Fourth and finally, there is a separate sense in which there is a meaningful global response to the pandemic, and in which this response may unfairly impose burdens on some-especially vulnerable populations such as refugees and people in low-income countries seeking better economic prospects. One of the approaches used to help address the crisis, which we have not focused on in this paper, is to sharply reduce international travel to isolate cases in particular countries and to help limit the spread of the disease. This policy has benefits to people in many countries. It does, however, disproportionately burden those who rely on the ability to travel across borders. The especially burdened members of this population include refugees seeking asylum, economic migrants, those seeking medical care in another country, and international students. In some cases, this population is being required to undertake a significant burden in order to benefit others who are not similarly burdened, and so there may be a separate reciprocity-based argument similar to the one we have given for individuals who benefit from travel restrictions to provide assistance to them.

The arguments in this paper have addressed the ethics of giving in response to the COVID-19 pandemic. We have argued, first, that such giving is not merely a matter of acting on one's duty of aid. This is because many of the devastating losses people are experiencing are not due to the coronavirus itself. Rather, some such devastating losses have also accompanied the collective efforts at mitigation. Such collective effort has involved significant sacrifices, especially those that accompany social distancing. Such sacrifices have produced various community-wide benefits such as reduced transmission rates and a flattened curve. Yet, while these benefits are vital, the burdens of the collective endeavor have not been borne equally. Indeed, the pandemic itself and the costs of social distancing have fallen disproportionately on less advantaged members of our society-in many cases, members of our society who are less advantaged due to historical and ongoing injustices. This is not merely tragic; it is profoundly unfair.

Our government has failed to take the requisite steps to ensure that citizens fulfill their obligations of reciprocity to those who have borne unfair burdens. Moreover, we have given reasons to think it will continue to fail in doing so. Accordingly, it falls to us to fulfill these obligations of reciprocity. Fully specifying how we should go about doing so will need to be sensitive to further empirical considerations that we do not have room to address. In the meantime, however, we have hopefully helped readers to think through how they should go about giving away money during the pandemic.

The authors would like to thank Brian Hutler, the audience at a workshop at the Berman Institute of Bioethics at John Hopkins University, and two anonymous referees for helpful comments on earlier drafts of this paper.

Pierce Randall https://orcid.org/0000-0002-0900-0985 ENDNOTES 1 See, for example, Maragakis (2020) , Virginia Department of Health (2020), and WBIR (2020). 2 A similar objection appears in the context of debates about whether black Americans are obligated to integrate neighborhoods so as to realize requirements of justice. Tommie Shelby argues that such requirements effectively impose additional burdens on individuals precisely because those individuals have already been subject to injustice. See Shelby (2014, 281-83) . 3 See Faden (2020); Lopez, Raine, and Bundiman (2020); Parker, Horowitz, and Brown (2020); and Solomon and Hamilton (2020) . 4 Our claim is not that everyone has direct, first-personal access to their own preferences and expectations to know whether or not they experience a burden. Rather, it is only the intuitive claim that people who do not believe that they are burdened by the collective response to COVID-19 are normally not being burdened. The converse does not follow: people who are not excessively burdened may nevertheless believe that they are excessively burdened, because they place greater weight than is warranted on minor inconveniences. Very wealthy investors, for example, may feel that they are unfairly burdened by a drop in the value of their portfolio. We claim that they are mistaken about that belief. While people sometimes mistakenly overestimate the degree to which they are burdened by social distancing, normally they are responsible for forming well-founded beliefs about their own condition relative to others. Someone who is not greatly burdened should not overestimate the importance of minor setbacks they experience, and so should donate government assistance they receive. Failure to do so would, at least in many cases, be a failure to discharge their duty of reciprocity because of culpable ignorance. 5 For a discussion of a similar kind of case, cf. Hume (2000, sect. 3.2.5) .

Stimulus Checks to Charity: 'It's Not Really My Money, It's Our Money

$200 Billion in Hazard Pay: These Workers Must Get Their Fair Share

Socioeconomic Disparities in Health in the United States: What the Patterns Tell Us

Just Health: Meeting Health Needs Fairly

A National and State Plan To End the Coronavirus Crisis

What's at Stake When Schools Close for the Coronavirus? A Bioethicist Weighs In

Voluntary Euthanasia and the Inalienable Right to Life

National Coronavirus Response: A Road Map to Reopening

The Coronavirus Burden is Falling Heavily on Black Americans. Why?" The Guardian

The Use of Knowledge in Society

A Treatise of Human Nature

Public Goods Theories of the Non-Profit Sector: Weisbrod Revisited

The Principle of Fairness and Political Obligation

Financial and Health Impacts of COVID-19 Vary Widely by Race and Ethnicity

Coronavirus, Social and Physical Distancing and Self-Quarantine

50 Million Kids Can't Attend School. What Happens to Them?

The Imperative of Personal Sacrifice, Today and During World War II

Failing Another National Stress Test on Health Disparities

About Half of Lower-Income Americans Report Household Job or Wage Loss Due to COVID-19

Legal Obligation and the Duty of Fair Play

A Theory of Justice

The Capability Approach

The Pure Theory of Public Expenditure

The Coronavirus Was an Emergency until Trump Found Out Who Was Dying

Integration, Inequality, and Imperatives of Justice: A Review Essay

Thinking About Racial Disparities in COVID-19 Impacts Through a Science-Informed, Early Childhood Lens

Famine, Affluence, and Morality

The Coronavirus Pandemic and the Racial Wealth Gap

Essential Workers are Taking Care of America. Are We Taking Care of Them?" Vox

Justice Without Borders: Cosmopolitanism, Nationalism, and Patriotism

Virginia Department of Health

Variation in COVID-19 Hospitalizations and Deaths Across New York City Boroughs

Stay Apart' Campaign Video

How to cite this article: Randall P, Bernstein J. Reciprocity and the ethics of giving during pandemics